Modulation of the IL-6 Receptor α Underlies GLI2-mediated regulation of Immunoglobulin Secretion in Waldenström macroglobulinemia cells

Cell culture and reagents

The IgM secreting cell line BCWM.1 (17, 18) was a kind gift from Dr. Steven Treon (Dana Farber Cancer Institute, Boston, MA). MWCL-1 cells (19) were a kind gift from Dr. Stephen Ansell (Mayo Clinic, Rochester, MN) and RPCI-WM1 cells (20) were kindly provided by Dr. Asher Chanan-Khan (Mayo Clinic, Jacksonville, FL). All cells were grown in RPMI-1640 supplemented with 10% FBS and penicillin/streptomycin. The GLI1/2 inhibitor (GANT61) and HH inhibitor (Cyclopamine) were obtained from EMD Millipore (Billerica, MA). The pan-caspase inhibitor (Q-VD-OPh) and all primers were obtained from Sigma-Aldrich (St. Louis, MO). β-actin antibody was obtained from Novus Biologicals (Littleton, CO). IL-6Rα antibody for western blot was from Santa Cruz Biotechnology (Dallas, TX). FITC-conjugated Annexin V was obtained from BD Biosciences (San Jose, CA) and propidium iodide was from Sigma-Aldrich.

Plasmid constructs and cell transfections

Short hairpin RNA (shRNA) targeting GLI1 and GLI2 were purchased from Origene Technologies, Inc (Rockville, MD) using plasmid vector pGFP-V-RS. The following shRNA targeting sequences were used: shGLI1 – TCTCACCTCTGTCGGATGCCAGCCTGGAC; shGLI2(1) – CCACGGAAAGCACTGGCTTCTCTGACAAC; shGLI2(2) – TCACCAAGAAGCAGCGCAATGACGTGCAC. The human IL-6Rα expression construct was kindly provided by Dr. Kishimoto (Osaka University, Osaka, Japan). The human IL-6Rα promoter (−853/+234 from first exon) was cloned into pGL3 basic using following primers: forward 5’-AGCTCAGGTACCCACGGCTCACTGTACCCTGGATCTCCCGG-3’ and reverse 5’-AGCTCAGCTAGCGGGTGCGGTTACGAAAGCCGGAGCGGATGGCGCAACAG-3’.

For all transfection experiments, 4 × 10⁶ cells were placed in cuvettes with plasmid DNA and electroporated at 250 V for 25 ms. Transfections were harvested after 48 h, unless otherwise noted. For knockdown experiments, 10 µg of shGLI1 or shGLI2 was used. For luciferase reporter experiments, 4 µg of IL-6Rα reporter was used. For the IL6-Rα rescue, 5 μg of IL6Rα and 10 μg of shGLI2 were used. For all experiments, empty vectors or scrambled non-specific shRNA were used as controls.

RNA Isolation and RT-PCR

Cells were collected and total RNA was isolated using TRIsure (Bioline, London, UK) per the manufacturer’s protocol. Reverse transcription (RT) reactions were conducted using M-MLV reverse transcriptase (Promega Corporation, Madison, WI). RT-PCR was performed using the EmeraldAmp GT PCR Master Mix (Clontech, Mountain View, CA) per the manufacturer’s protocol. PCR products of were analyzed by gel electrophoresis using the following forward and reverse primers: GLI1 – ACTGAAGACCTCTCCAGC (forward) and GCTGACAGTATAGGCAGA (reverse); GLI2 – GACCACAGCATGCTCTACTACTA (forward) and ATGTTCAGGAACTTGCTCTCCTCG (reverse); GLI3 – CCAACGGAAATCAATAGGAGTTGAA (forward) and GAGGGTGGTTTGAGTGTAACAATAC (reverse); BAFFR – ACAAAGGTCATCATTCTGTCTCC (forward) and CCTTCAAGGGCTGTCAAAGA (reverse); BCMA – CCTGTTTGGGACTGAGCTTAATA (forward) and AGCTGTATCGGCAACTGAAA (reverse); IL-2Rα – GTGAGACTTCCTGCCTCGTC (forward) and GGGACTGAGCTGGCATAGAG (reverse); IL-6Rα – CTCAGTGTCACCTGGCAAGA (forward) and TGTCGCATTTGCAGAATCTC (reverse); IL-21R – CTCTGACTGGGTCTGAAAGATG (forward) and GACCAGTATGAAGAGCTGAAGG (reverse); TACI - CAAGGAGCAAGGCAAGTTCTA (forward) and CCTCTGTCTCTCTCTCCTCATATC (reverse).

Real-time PCR (qPCR) was performed using ViiA 7 Real-Time PCR System by Life Technologies (Grand Island, NY). For copy number analysis, 2.0 × 10⁶ cells of BCWM.1, MWCL-1 and RPCI-WM1 were grown in 3.0 mL of RPMI-1640 (+10% FBS & P/S) for 48 hr. Copy number of Gli1, Gli2, and Gli3 was determined in each cell line using the following primers: GLI1 – GCTACATCAACTCCGGCCAAT (forward) and CGGCGTTCAAGAGAGACTGG (reverse); GLI2 – CACGTCAGAGCCATCAAGACC (forward) and ACTGTCATTGTTGGGGGCACT (reverse); and GLI3 – GAATGCCTCCAGGTGAAAACC (forward) and AGTTGCTGATGGGGGACTGTT (reverse). A standard curve was generated using a plasmid with the coding region of the gene of interest incorporated into it. For expression of IL-6Rα relative to the expression of the housekeeping gene, GAPDH, the following primers were used: IL-6Rα – GTGCTCTTGGTGAGGAAGTT (forward) and TTCTGGGACTCCTGGGAATA (reverse); GAPDH – CTCGACTTCAACAGCGACA (forward) and GTAGCCAAATTCGTTGTCATACC (reverse). For amplification of mouse IL-6Rα, the following primers were used for qPCR: GAPDH – CGTCCCGTAGACAAAATGGT (forward) and TTGATGGCAACAATCTCCAC (reverse); and IL-6Rα – GGCTCACAAAACAGAGAATGG (forward) and GGTGATCATTCAGGGAGCAT (reverse).

Enzyme-linked Immunosorbent Assay (ELISA)

ELISA plates (Nunc maxisorp; Nalgene Nunc International, Rochester, NY) were used to quantitate IgM levels. Human and mouse IgM ELISA (Bethyl laboratories, Inc, Montgomery, TX) were used following manufacturer recommendations. Reactions were stopped by the addition of 1N H₂SO₄ and the results were measured with an Epoch plate reader (BioTek, Venooski, VT).

Imunoblotting

Total protein lysate was collected from BCWM.1, MWCL-1, and RPCI-WM1 cells and protein concentration was determined using a BCA protein assay kit from Thermo Fisher Scientific, Inc. (Waltham, MA). Protein (100 μg) was loaded into each well of a 7.5% SDS protein gel and a semi dry transfer system (BioRad, Hercules, CA) was used to transfer the protein to a nitrocellulose membrane. Antibody specific for IL-6Rα or β-actin was used to detect protein and β-actin antibody was used as a reference. Densitometry was performed using ImageJ software, normalized to control and presented as an average of 3 independent experiments.

Cell viability

Cells (0.25 × 10⁶) treated with GANT61 or Cyclopamine were incubated in 48-well plates in RPMI-1640 containing 10% FBS for 48 h. Cells were washed with Annexin V binding buffer and stained with 1 μg Annexin V FITC (BD Biosciences) for 20 min at 4°C in the dark. Cells were again washed with Annexin V binding buffer and then stained with 0.5 μg propidium iodide. Cells were immediately processed using a FACSCalibur (BD Biosciences) and data was analyzed using FlowJo software (FlowJo LLC).

Proliferation assay

Cells were suspended in RPMI-1640 (without phenol red, 5% FBS, + penicillin/streptomycin) at a concentration of 0.25 × 10⁶ cells/mL. 100 μL of the cell suspensions were plated in triplicate in a 96-well plate and treated with 20 μM GANT61 or DMSO as a control. Final volume was brought to 200 μL/well. After three days, 50 μL of XTT working solution (Trevegin, Gaithersburg, MD) was added to each well and incubated at 37°C for 1 h and then analyzed on an Epoch plate reader (Biotek).

Luciferase reporter assay

Cells (4.0 × 10⁶) were transfected, as described above, plated in triplicate in 24-well plates, and incubated in RPMI-1640 supplemented with 10% FBS. After 48 h, cells were harvested and assayed for luciferase activity per the manufacturer’s protocol (Promega, Madison, WI). In order to control for variations due to transfection efficiency, total protein for each sample was quantified using a BCA protein assay kit (Thermo Fisher Scientific, Inc). Luciferase readouts were normalized to total protein content and relative luciferase represents luciferase readouts/protein concentration relative to control samples in each experiment.

Chromatin immunoprecipitation (ChIP) assay

ChIP was conducted using a modification of the Magna ChIP kit protocol (EMD Millipore, Bedford, MA) as previously described (21). Chromatin was digested with micrococcal nuclease (2 gel units/µl; New England Biolabs, Ipswich, MA) for 20 min at 37°C. Chromatin was further sheared by sonication for 15 cycles of 30 sec on/30 sec off at 4°C (Bioruptor, Diagenode, Denville, NJ). Aliquots of the sheared chromatin were subjected to immunoprecipitation using anti-GLI2 antibody (Novus Biologicals) or normal rabbit IgG (EMD Millipore). Real-time PCR of the ChIP products and genomic input DNA was performed using primers that amplify an area of the IL-6Rα promoter containing consensus GLI binding sites. The sequences of the primers are the following: CCAATCACGGCTCACTGTA (forward) and TTACAGGCGTGAGCCAC (reverse). Quantitative SYBR PCR was performed at 59°C in triplicate for each sample or control using the C1000 Thermal Cycler (Bio-Rad) and the results presented as percentage of input.

Peritoneal cell isolation and culture

Normal B cells were isolated from female Balb/cByJ mice (5-8 weeks) that were obtained from the Jackson laboratory (Bar Harbor, ME). Mice were cared for and handled in accordance with Institutional and National Institutes of Health guidelines. Peritoneal washout cells were obtained as previously described (22). B-1a cells were enriched by a B220 negative selection process followed by CD5 positive selection using mAb-coupled MicroBeads (Miltenyi Biotec, Auburn, CA) following manufacturer’s instructions. Purity was determined ≥95% by FACS. Total peritoneal cells or purified B-1a cells were cultured at 1*10⁶ cells/ml in 6 or 12-well plates in the presence or absence of the indicated treatments.

Flow cytometry

Fluorochrome-conjugated antibodies used for flow cytometry were obtained from BD Biosciences. For GLI detection, 0.5*10⁶ WM B cells were collected in FACS tubes. The cells were then permeabilized and stained with the BD Cytofix/Cytoperm Plus kit (BDBiosciences 555028) according to manufacturer recommendations with some modifications. Briefly, cells were spun down and resuspended in 100 µl of Fix/Perm solution and incubated at 4°C for 30 min. Cells were then washed twice with Fix/Perm Wash Buffer and resuspended in 50 ul Fix/Perm solution. Four µg of primary antibody from Novus for GLI1 (NB600-600), GLI2 (NB600-874), GLI3 (NBP2-29627) and Isotype control (NBP2-36463) were added and the tubes were incubated at 4°C for 1h. Cells were washed twice with Fix/Perm Wash buffer and resuspended in 50 µl Fix/Perm solution. Anti-Rabbit IgG (PE Conjugated; 10 µl) secondary antibody from BD (558416) was added. Tubes were incubated at 4°C for 45 min and then washed twice with Fix/Perm Wash Buffer. Cells were resuspended in 500 µl FACS Buffer and analyzed via Flow-Cytometry. Cells were analyzed on a FACSCalibur instrument and analyzed using FlowJo software.

Statistical analysis

A two-tailed Student’s t-test was used to determine statistical significance between data 2 sets or ANOVA between >2 data sets. A p<0.05 was considered significant and figures denoted with ‘*’ indicate p<0.05 and ‘**’ indicate p<0.0001. Analysis was carried out using Prism software (GraphPad, La Jolla, CA).

GLI inhibition reduces IgM secretion

Because of the oncogenic role of GLI proteins described in several malignancies, we initially screened for the expression of GLI family members and found that WM cell lines, MWCL-1 and RPCI-WM1 and the IgM secreting cell line BCWM.1, express GLI1, GLI2 and GLI3 (Figure 1A and 1B). To determine if GLI played a role in WM biology, we used the GLI pharmacological inhibitor GANT61 and found a significant reduction in IgM secretion in response to GANT61 treatment in BCWM.1 (p<0.0001), MWCL-1 (p<0.0001) and RPCI-WM1 (p<0.0001) cells (Figure 2A). As previously described (23), GANT61 treatment reduced the expression of GLI1 and GLI2 (Figure 2B) in our model cells. Investigation of the effect of GLI inhibition on cell growth and survival indicated GLI inhibition had a modest effect on the proliferation of BCWM.1 (20% reduction) (p=0.0032) and RPCI-WM1 (15% reduction) (p=0.0381) cells while it had no effect on MWCL-1 cells (Figure 2C). GLI inhibition also resulted in a modest (14%) reduction in cell viability in BCWM.1 (p<0.0001) and 15% in MWCL-1 (p<0.01) (Figure 2D). This modest reduction in growth and survival is likely not responsible for the significant reduction in IgM secretion observed in cells treated with GANT61, as pretreatment of cells with the pan-caspase inhibitor Q-VD-OPh (QVD) had no effect on IgM secretion by cells treated with GANT61 (Figure 2E) although it did restore cell viability lost by GLI inhibition in MWCL-1 cells (p<0.0001)(Figure 2F). Taken together, these results suggest that GLI inhibition results in a reduction in IgM secretion that is independent of cell viability.

Hedgehog (HH)-independent regulation of IgM

GLI transcription factors were initially described as the effectors of the Hedgehog (HH) signaling pathway (24, 25). Therefore, we screened for the expression of the HH receptors patched 1 (PTCH1) and smoothened (SMO). Using PCR, we found that WM cells express PTCH and SMO (Figure 3A) suggesting HH signaling may play a role in GLI-mediated regulation of IgM secretion. Using the pharmacological inhibitor of HH signaling Cyclopamine, we found that cells treated with this inhibitor had no change in IgM secretion (Figure 3B). Expression of GLI1, a marker of the activity of the HH pathway, was used to confirm the effect of Cyclopamine in our cell line models. As expected, the inhibitor lowered the levels of GLI1, demonstrating the efficacy of the treatment (Figure 3B). Interestingly, Cyclopamine treatment had no effect on GLI2 or GLI3 expression in these cells (Supplemental Figure 1A). Further investigation indicated that BCWM.1 cells treated with cylopamine had a reduction in cell growth (p=0.0055) (Figure 3C) and cell viability (p=0.0007) (Figure 3D), with no change in MWCL-1 or RPCI-WM1 cells. Taken together, these results suggest that signaling mediated through GLI but not SMO is likely responsible for the regulation of IgM secretion.

IL-6Rα is a direct target of GLI2

To characterize the mechanism by which GLI inhibition reduces IgM secretion, we hypothesized that GLI transcription factors may modulate expression of cytokines or cytokine receptor genes known to modulate IgM secretion (26-33). We therefore screened for the expression of receptors for IL-2, IL-6, IL-21 and B-lymphocyte stimulator (BAFF). We found that GLI inhibition using GANT61 reduced IL-6Rα mRNA expression, while having no effect on other cytokine receptors investigated in MWCL-1 cells (Figure 4A). We confirmed these findings using qPCR in RPCI-WM1 cells, where we found that GLI inhibition significantly reduced IL-6Rα expression (p<0.0001) cells (Figure 4B). Because GANT61 reduces GLI1 and GLI2 mRNA expression (13, 23, 34) and Figure 2B, we next targeted GLI1 and GLI2 using RNAi to determine the involvement of each in regulating IL-6Rα. We found that GLI2 knockdown but not GLI1 reduced IL-6Rα mRNA (Figure 4C) and protein expression (Figure 4D). Consistent with expression data, GLI2 knockdown reduced the surface expression of IL-6Rα (Figure 4E). Finally, to confirm the specificity of the RNAi targeting, we used 2 different shRNA sequences targeting GLI2, and found a reduction in IL-6Rα expression and IgM secretion (Figure 4F and Supplemental Figure 1B). A control showing a reduction in GLI2 protein expression was also performed in all 3 cell lines (Supplemental Figure 1C). Taken together, these experiments show that GLI inhibition reduces IL-6Rα expression and that GLI2, but not GLI1, reduces IL-6Rα expression and ultimately IgM secretion.

Using bioinformatics analysis, we found several candidate GLI binding sites on the IL-6Rα promoter region suggesting GLI2 may modulate IL-6Rα expression by modulating its promoter (Figure 5A). We therefore cloned the IL-6Rα promoter into a luciferase reporter construct and used that to determine the ability of GLI2 to modulate this regulatory sequence. We initially tested this reporter construct in our cell lines compared with empty vector (Supplemental Figure 2). Consistent with our expression data, GANT61 treatment significantly reduces IL-6Rα promoter activity in BCWM.1 (p=0.0027), MWCL-1 (p=0.0017) and RPCI-WM1 (p=0.0004) cells (Figure 5B). Furthermore, GLI2 knockdown significantly decreased IL-6Rα promoter activity (p<0.0001 for each cell line) (Figure 5C). Using chromatin immunoprecipitation (ChIP) assay followed by quantitative PCR (ChIP-PCR), we found direct binding to the IL-6Rα promoter in the distal region in BCWM.1, MWCL-1 and RPCI-WM1 cells (Figure 5D). Next we performed ChIP assay on MWCL-1 cells treated with either GANT61 or vehicle control and found a significant reduction in GLI2 binding to the IL-6Rα promoter in cells treated with the GLI inhibitor (p=.0174)(Figure 5E). Similarly, in cells transfected with shGLI2, there was a significant reduction in GLI2 binding to the IL-6Rα promoter compared with control cells transfected with the control Scramble shRNA (p<.0001) (Figure 5F).

Finally, to confirm the biological role of GLI2- IL-6Rα axis, we overexpressed IL-6Rα in cells transfect with shRNA targeting GLI2 and determined the effect on IgM secretion by ELISA. We found that overexpression of IL-6Rα can rescue the reduction in IgM secretion that is observed with GLI2 knockdown (p<0.0001) (Figure 6A). This data suggest that GLI2 modulates IgM secretion by regulating an important cytokine receptor, the IL-6Rα.

GLI2-IL-6Rα-IgM axis is active in natural IgM-secreting cells

Our data so far was performed using B-cell lines derived from patients with WM. To confirm the role of GLI2 in modulating IL-6Rα and IgM secretion in primary cells, we used mouse B-1 cells, which secrete natural IgM. Because these cells exist in high numbers in the peritoneal cavity (35, 36), we treated total peritoneal cells with or without GANT61. After 2 days of culture, cells were harvested and stained with B220 and CD23 to distinguish B-1/B-2 cells (22). We found that B-1 cells treated with GANT61 had reduced surface expression of IL-6Rα (Figure 7A). We also found a significant reduction in IgM secretion in culture supernatant harvested from cells treated with GANT61 (p<0.0001)(Figure 7B). Next, we purified B-1a cells and used them to determine the effect of GANT61 on IgM secretion. Similar to the results obtained with total peritoneal cells, treatment of B-1a cells with GANT61 results in a significant reduction in natural IgM in culture supernatants by ELISA (p<0.0001)(Figure 7C). We next confirmed the expression of GLI1, GLI2 and GLI3 in B-1a cells by FACS analysis (Figure 7D). Finally, we found that treatment of B-1a cells with the GLI inhibitor GANT61 reduces IL-6Rα mRNA expression (p=0.0030)(Figure 7E) and GLI2 and IL-6Rα protein expression (Figure 7F). Taken together, these data confirm that GLI2 regulatory mechanism controlling IL-6Rα and ultimately IgM secretion is conserved beyond WM cells and may have important physiological functions under non-disease conditions.